Inland penetration of wintertime maritime showers

Convective precipitation produced by the IFS convection scheme is considered to fall out immediately at the grid point.  It is not advected laterally irrespective of the winds experienced during descent of the precipitation.  Active maritime convective cells can persist for an hour or so and therefore in reality can reach well inland.  In Cy49r1 and earlier, this resulted in a systematic bias towards too little inland penetration of coastal showers. 

In contrast, precipitation produced by the IFS broad scale precipitation scheme is advected by the winds at each level.   Precipitation reaching the surface can be well away from the location where it was produced, even across several grid squares.

A modification introduced in Cy50r1 improves the distribution of forecast convective precipitation, particularly inland penetration.  In this scheme, a proportion of the convective precipitation is passed to the broad scale scheme.  The precipitation can then advect according to the model winds during its descent and make progress inland.  Note that snow flakes advect further in the wind than rain because of slower fall speed. Comparing Cy50r1 with Cy49 and earlier:

  • Cy50r1 has greater areal extent of precipitation (rain or snow) over the sea in the convective westerlies, and with land convection in the warmer seasons.  Cy49r1 has less extent. of precipitation.
  • Cy50r1 has more variation in the distribution of precipitation (rain or snow) rates.  Cy49r1 has a flatter distribution of rates - more zeros and higher peaks.
  • Cy50r1 seems far better with snow showers extending downwind, including inland from coasts.  Cy49r1 caused all precipitation to fall to the surface immediately.
  • Cy50r1 brings likelihood that meteograms of precipitation type will have higher bars (i.e. higher probabilities) and precipitation type ENS-based map charts will be more colourful because of these higher probs.

See also section on convective cloud processes and precipitation.

Example of inland penetration of showers

Fig9.6.1-1: An example of deeper penetration inland of active lake showers in Cy50r1.   DT00UTC 30 Dec 2025, VT T+48 00UTC 01 Jan 2026.  Evidence of greater downstream extension of lake effect snow in 50r1 (to the ESE).  Also greater upstream extension in 50r1 (to the WNW), and slightly larger amounts overall.  Snow over sea can be greater in 50r1, mostly affected by the changes in evaporation and the increase in the fall speed of snow in the cloud scheme.   Advise on what the snow fall speed change is (something to include for users in the PR material for 50r1) See FDR 31/12/25.



  

Effect of onshore wind strength (Cy49r1 and earlier)

In Cy49r1 ands earlier, any showers that IFS develops over the sea do not penetrate beyond the coast.  Inland showers are developed using the local model atmospheric structure that has moved in from upwind.  Therefore shower precipitation is biased rather too much in coastal parts, rather too little inland.  The areas affected are extended with stronger flow.

Fig9.6.1-5: Schematic illustration of systematic precipitation biases in onshore maritime convection.  Too much precipitation is forecast for windward coastal zones.  Too little precipitation is forecast for areas leeward of high ground.  These areas expand and move downwind with stronger winds. 

In strong winds, maritime showers could extend further inland than Ensemble Control Forecast showed because in reality:

  • active maritime convective cells can persist for an hour or so.
  • rain can be advected laterally during descent.  
  • snow flakes drifts further than rain because of slower fall speed.

Cy50r1 (introduced May 2026) has helped to reduce these limitations in the convective products.





(FUG associated with Cy50r1)





  • No labels